| In a time of tremendous biomedical research activity linked to both the sequencing of the human genome and the availability of high-throughput technologies that measure functional aspects of the cell, it is important to foster the conception and implementation of methods that integrate the avalanche of new research data. In this thesis, we explore the use of text mining for automatically capturing molecular information from a rapidly expanding pool of scientific articles. We discuss methods of building intelligent tools to extracte biological facts from the literature, dates management issues related to large-scale text mining and the use of text mining to answer real-world biological questions. By using a literature-compiled molecular interaction network for the prediction of toxic drug affects, we demonstrate that the automated collection and integration of published, readily available biological information is a powerful method for tasting biomedical hypotheses. |
